Impact of endocervical specimen quality on apparent prevalence of Chlamydia trachomatis infections diagnosed using an enzyme-linked immunosorbent assay method.

Duplicate endocervical swabs were collected from 1824 patients for detection of Chlamydia trachomatis. Specimen pairs were combined into 400 microL of 0.9% saline solution. After vortexing, a 40-microL sample was smeared and stained with Papanicolaou's method for detection of endocervical and/or metaplastic (E-M) cells. The remaining specimen was tested for C trachomatis antigen with the use of an enzyme-linked immunosorbent assay (ELISA) procedure (Chlamydiazyme, Abbott Laboratories, North Chicago, Ill). Chlamydia trachomatis antigen was detected and confirmed (with the use of a blocking antibody [Abbott Laboratories]) in only 16 (1.7%) of 918 specimens that lacked detectable E-M cells, but it was detected significantly more frequently not only in 88 (13.3%) of 661 specimens that contained detectable E-M cells but also in 32 (13.1%) of 245 specimens that contained too many red blood cells to analyze microscopically. Of the initially positive ELISA results, none of 37 were falsely positive from specimens that contained 11 or more E-M cells, but significantly more (six [27.3%] of 22) were falsely positive from specimens that lacked detectable E-M cells. Variations in specimen quality had a significant impact on the incidence of both true-positive and false-positive ELISA results and could significantly influence understanding of the prevalence of chlamydial infections in women.